AI-Assisted Development of a C++ RTS Game in Unreal Engine 5

Abstract

This thesis introduces a real-time strategy (RTS) game prototype inspired by the Battle of Hanzhong from Romance of the Three Kingdoms. The prototype features several battlefield elements, including mountain stronghold assaults, ambush tactics, and open-field engagements between the Shu and Wei forces. It is hoped to create an RTS with similar ease of operation as World of Warcraft, featuring the charging of generals, the compatibility of different military units, and the influence of morale on combat strength, etc.

The project followed an iterative workflow and received continuous support from Professor Jukka. Based on his feedback, Unreal Engine was chosen over custom frameworks, and detailed documentation was maintained with the help of AI and GitHub. Unreal Engine ultimately provided a suitable environment for building a large 3D RTS map while allowing continued practice with C++ programming.

Under the guidance of Instructor Teemu, the academic focus of the thesis was finalized to examine how AI supports the development process. The study looks at how AI assisted the creation of 3D maps in Unreal Engine, contributed to game performance improvements, and helped analyze C++ code structures. The findings show that AI can meaningfully enhance workflow efficiency and provide technical support in complex RTS game development.